The pervasive use of Online Social Networks (OSN) for networking, communication and search in tandem with the ubiquitous availability of smartphones, which enables real-time multimedia capturing and sharing, have led to massive amounts of user-generated content and activities being amassed online, and made publicly available for analysis and mining. Each content item is associated with an abundance of metadata and related information such as location, tags, comments, favorites and mood indicators, access logs, and so on. At the same time, all this information is implicitly or explicitly interconnected based on various properties such as social links among users, groups, communities, and sharing patterns. These properties transform social media into data sources of an extremely dynamic nature that reflect topics of interests, events, and the evolution of community opinion and focus. Social media processing offers a unique opportunity to structure and extract information and to benefit multiple areas ranging from new media experiences to psychology and marketing. The objective of this talk is to provide an overview of the current research in emerging topics related to applications where social media can act as sensors of real-life phenomena and case studies that reveal valuable insights. After discussing challenges and presenting a generic conceptual architecture, there will be a focus on efficient processing and indexing algorithms that can handle massive amounts of content with application to graph-based event detection and summarization in social media streams.